Predicting Ergodic Data Capacity and BER for MIMO Communication Systems using SVM Model

Abstract

The maximum error-free data rate that a channel can support is known as the data handling capacity of the channel. In today's era of widespread digital communication, all channels inevitably aim to operate at their maximum data rate. Over the years, researchers have extensively explored various technologies and modulation techniques to boost data transmission speeds. Orthogonal frequency division multiplexing followed by a multiple input multiple output system is used for 5G technology to achieve an enhanced data rate. However, as is true for all systems and designs, MIMO systems also have pockets of vulnerability. The two most sensitive parameters are the impairment level at the transmitter end and a variable correlation coefficient at the receiver end. In this paper, we have concentrated on the receiver end distortion because it affects the data rate capacity. In MIMO systems, at the receiver end, the kappa factor, which is a function of the channel impulse response, and α, a factor dependent on the correlation coefficient of the identical antennas, significantly influence the maximum ergodic data rate and the corresponding bit error rate. In this work, we have applied the support vector regression model to simulate the non-linear nature of these issues and established a relation between the SVM-predicted result and data obtained from an opensource data set. The proposed scheme helps the designers and propagation engineers derive realistic data rate values by considering two important factors.